This paper studies the loss of efficiency of polymeric drag reducers induced by high Reynolds number flows in tubes. The overall pressure was fixed and the apparatus was built so as to minimize the polymer degradation. We used three kinds of polymers: two flexible and one rigid. We conducted our tests to take into account the drag reduction (DR) for a wide range of concentrations of each polymer. The main results are displayed for the DR as a function of the number of passes through the apparatus. The mechanism of the loss of efficiency for the Xanthan Gum (XG) solutions (the rigid one) seems to be completely different from that observed for Poly (ethylene oxide) (PEO) and Polyacrylamide (PAM) (the flexible materials). While the PEO and PAM mechanically degrade by the action of the turbulent flow, the XG seems to remain intact, even after many passes through the pipe flow apparatus. From the practical point of view, it is worth noting that the PAM solutions are clearly more efficient than the PEO and XG. Another practical point that deserves attention is concerned with the asymptotic drag reduction found for XG. Although its maximum DR was significantly smaller than that found for PEO, the final value for both polymers were quite the same, which is obviously related to the intensified mechanical molecule scission in the PEO solutions. Our results for the relative drag reduction (the current value of DR divided by its maximum obtained at the first pass) was quite well fitted by the decay function proposed in our previous paper [A. S. Pereira and E. J. Soares, “Polymer degradation of dilute solutions in turbulent drag reducing flows in a cylindrical double gap rheometer device,” J. Non-Newtonian Fluid Mech. 179, 9–22 (2012)], in which a rotating apparatus was used. This strongly suggests that the physical mechanism that governs the degradation phenomenon is independent of the geometry. We also used a degradation model for PEO proposed by Vonlanthen and Monkewitz [“Grid turbulence in dilute polymer solution: Peo in water,” J. Fluid Mech. 730, 76–98 (2013)] to fit our data of relative drag reduction for PEO and PAM.

The polymer coil–stretch mechanism in turbulent drag reducing flows is analysed using direct numerical simulations of viscoelastic finitely extensible nonlinear elastic fluids with the Peterlin approximation. The study is carried out taking into account low and high drag reduction regimes. The polymer stretching and the alignment between the conformation tensor and other relevant entities are investigated using statistical and tensor analysis. The significant alignment between the former and the velocity fluctuations product tensor indicates that the initial polymer stretching due to the mean shear is increased by the flow stress fluctuations, providing a supplementary polymer extension. In addition, interactions between the turbulence and the polymer are evaluated from an instantaneous turbulent energy exchange perspective by considering streamwise work fluctuating terms in elliptical and hyperbolic flow regions separately. Near the wall, polymers not only release energy to the streaks, but also to the elliptical (or vortical) and hyperbolic (or extensional) structures. However, polymers can also be dragged around near-wall vortices, passing through hyperbolic regions and experiencing a significant straining within both these turbulent structures and storing their energy. Hence, polymers weaken elliptical and hyperbolic structures leading to a tendency toward relaminarization of the flow. Polymer release of energy occurs primarily in the streamwise direction, which is in agreement with the enhanced streamwise velocity fluctuation observed in drag reducing flows. A detailed polymer coil–stretch mechanism is provided.